(Responding to this (Arxiv pdf) link to a Sean Carroll think piece on cosmology and the multiverse; for the sci-fi writers in the group the multiverse is a standard part of our toolkit now; for the science branch of my world, it's sort of a philosophical kludge, the mathematics of the thing are undeniable in what's going on, it's our attempt to turn that mathematics into a physical reasoning that reaches a limit in translation at the current epoch)

Ok, Sean, stipulated: any physical/mathematical theory (equation, in this context) that seeks to describe an observable spacetime event is scientific. I've got two questions, though.

One is the problem that all hidden variables share: they're ultimately the same thing as having an extra set of parameters to fit a curve. A theory which can fit a curve with three parameters, say, owes little to a theory which has forty-two parameters (to exaggerate and pun simultaneously), except only for the case where an interesting new mathematics can be developed from the different descriptions (this is a significant thing to theoreticians, it's essential for coming up with new methods e.g. Einstein, Heisenberg, Dirac, Feynman etc).

But my second question: neither case, in my view, actually describes what's going on with multiverse in the colloquial sense versus the mathematical-physical description that might ultimately lead us somewhere. In precise terms, we mean that the (human) measurable universe has a conceptual lightspace volume, beyond which we have no current means of accessing. However, the numbers we use as basis for the equations involved extend well past our physics description's current restrictions for that lightspace volume (meaning, regardless of our explanation for why, precisely, our telescopes can still only see so far into the darkness).

In which case, the real meat here, from a theoretician's point of view, is whether, if we pick any given means for reconciling the countably infinite number line with the finite limit of observable spacetime in our current physical equations, do we get anything new and interesting (i.e. predictive) out of it?